[Git][freetype/freetype][master] [smooth] Reduce shift in multiply-shift optimization.

[Alexei Podtelezhnikov (@apodtele)]

Alexei Podtelezhnikov pushed to branch master at FreeType / FreeType

Commits:

• b6c11d49
  by Alexei Podtelezhnikov at 2021-08-30T07:09:53-04:00

  [smooth] Reduce shift in multiply-shift optimization.
  
  * src/smooth/ftgrays.c (FT_UDIVPREP, FT_UDIV): Reduce shift.
  
  Smaller shifts that keep the division operands of FT_UDIVPREP within
  32 bits result in slightly faster divisions, which is noticeable in
  the overall performance.  The loss of precision is tolerable until the
  divisors (the components dx and dy) approach 32 - PIXEL_BITS. With
  PIXEL_BITS = 8, this corresponds to 65,000 pixels or the bitmap size
  that we refuse to render anyway.
  
  Using `ftbench -p -s60 -t5 -bc timesi.ttf`,
  
  Before: 8.52 us/op
  After:  8.32 us/op
  

1 changed file:

• src/smooth/ftgrays.c

Changes:

src/smooth/ftgrays.c

---

...	...	@@ -386,12 +386,10 @@ typedef ptrdiff_t FT_PtrDist;
386	386	/* divisors to provide sufficient accuracy of the multiply-shift. */
387	387	/* It should not exceed (64 - PIXEL_BITS) to prevent overflowing and */
388	388	/* leave enough room for 64-bit unsigned multiplication however. */
389		-#define FT_UDIVPREP( c, b ) \
390		- FT_Int64 b ## _r = c ? (FT_Int64)( ~(FT_UInt64)0 >> PIXEL_BITS ) / ( b ) \
391		- : 0
392		-#define FT_UDIV( a, b ) \
393		- (TCoord)( ( (FT_UInt64)( a ) * (FT_UInt64)( b ## _r ) ) >> \
394		- ( sizeof( FT_UInt64 ) * FT_CHAR_BIT - PIXEL_BITS ) )
	389	+#define FT_UDIVPREP( c, b ) \
	390	+ FT_Int64 b ## _r = c ? (FT_Int64)0xFFFFFFFF / ( b ) : 0
	391	+#define FT_UDIV( a, b ) \
	392	+ (TCoord)( ( (FT_UInt64)( a ) * (FT_UInt64)( b ## _r ) ) >> 32 )
395	393
396	394
397	395	/* Scale area and apply fill rule to calculate the coverage byte. */

—
View it on GitLab.
You're receiving this email because of your account on gitlab.freedesktop.org. If you'd like to receive fewer emails, you can adjust your notification settings.